Each job scheduler in large decentralized load balancing systems gener
ally must consider whether it is advantageous to offload jobs to remot
e computation servers when the local load is too high. Although proces
sing power may appear to be available at a very distant server, two pr
oblems arise due to the transmission delay between the scheduler and s
erver. Predictably, the response time of the job is adversely affected
as the job spends valuable time in transit, but a more subtle problem
involves the value, or reliability, of the state information regardin
g job queues. The longer the delay between scheduler and server, the l
ess a scheduler should value the state information of the server (give
n that the state changes over time). We examine the performance of sch
edulers in topologies with different average proximity and show a prob
abilistic algorithm that allows schedulers to dynamically form efficie
nt clusters in the network.